Language disorders resulting from focal brain damage can take a variety of forms, and there is evidence that some aphasic deficits selectively affect the comprehension and interpretation of sentences. This project investigates the functional bases of the sentence processing disorders that occur in aphasia. The specific aims for the next project period involve continued investigation of the contribution of impairments of lexical/semantic representations and/or processes to symptoms involving sentence production and comprehension. Aphasic patients and normal control subjects will participate in a variety of different types of experiments: (1) The contribution of various types of information to lexical retrieval in sentence production will be investigated using tasks combining semantic, syntactic, visual/pictorial, or phonological cues to trigger word retrieval in timed production tasks. These studies will evaluate the relative contribution of these information sources to retrieval of nouns and verbs that are high and low in imageability, and will assess the degree to which competing responses interfere with retrieval. (2) New techniques will be designed to elicit sentences of different structural types from patients by exploiting the factors believed to affect lexical and structural decisions in normal sentence production; the relative effectiveness of these factors will be assessed in a separate set of experiments. (3) The contribution of semantic factors to impairment in the comprehension of reversible sentences will be studied using standard sentence/picture matching and speeded verification. The consequences of verb meanings for thematic role assignment are hypothesized to have specific and measurable effects on patients' comprehension performance. (4) Patients' sensitivity to violations of sentence meaning that are caused by illegal combinations of lexical semantic and structural information will be evaluated using the word monitoring task. Sentence complexity (passive v. active) is expected to attenuate violation detection for patients with poor sentence comprehension. The long-term goal is to develop a model sentence comprehension and production that can provide a functional explanation for the symptoms found in aphasia. Such a model would have a variety of applications. If specific aphasic symptoms can be interpreted as arising from identifiable processing deficits within a testable model, then attempts to remediate symptoms can be more definitely focused on the responsible representation or process. Moreover, discrete components of language processing that are identified through studies of aphasic patients are likely to be the components that will, ultimately, prove to be localizable in the brain.